修改tutorial中的typo; 增加DataBundle的add_collect_fn

docs/source/tutorials/序列标注.rst

@@ -147,6 +147,8 @@ fastNLP的数据载入主要是由Loader与Pipe两个基类衔接完成的，您
 .. code-block:: python
 
     from fastNLP.io import WeiboNERPipe
+    from fastNLP.models import BiLSTMCRF
+
     data_bundle = WeiboNERPipe().process_from_file()
     data_bundle.rename_field('chars', 'words')
 

docs/source/tutorials/文本分类.rst

@@ -301,7 +301,7 @@ fastNLP提供了Trainer对象来组织训练过程，包括完成loss计算(所
     
     # 这里为了演示一下效果，所以默认Bert不更新权重
     bert_embed = BertEmbedding(char_vocab, model_dir_or_name='cn', auto_truncate=True, requires_grad=False)
-    model = BiLSTMMaxPoolCls(bert_embed, len(data_bundle.get_vocab('target')), )
+    model = BiLSTMMaxPoolCls(bert_embed, len(data_bundle.get_vocab('target')))
     
     
     import torch

fastNLP/io/data_bundle.py

@@ -6,7 +6,7 @@ __all__ = [
     'DataBundle',
 ]
 
-from typing import Union
+from typing import Union, List
 
 from ..core.dataset import DataSet
 from ..core.vocabulary import Vocabulary
@@ -274,6 +274,22 @@ class DataBundle:
         for name, dataset in self.datasets.items():
             yield name, dataset
 
+    def get_dataset_names(self) -> List[str]:
+        """
+        返回DataBundle中DataSet的名称
+
+        :return:
+        """
+        return list(self.datasets.keys())
+
+    def get_vocab_names(self)->List[str]:
+        """
+        返回DataBundle中Vocabulary的名称
+
+        :return:
+        """
+        return list(self.vocabs.keys())
+
     def iter_vocabs(self) -> Union[str, Vocabulary]:
         """
         迭代data_bundle中的DataSet
@@ -332,6 +348,27 @@ class DataBundle:
             dataset.apply(func, new_field_name=new_field_name, **kwargs)
         return self
 
+    def add_collect_fn(self, fn, name=None):
+        """
+        向所有DataSet增加collect_fn, collect_fn详见 :class:`~fastNLP.DataSet` 中相关说明.
+
+        :param callable fn:
+        :param name:
+        :return:
+        """
+        for _, dataset in self.datasets.items():
+            dataset.add_collect_fn(fn=fn, name=name)
+
+    def delete_collect_fn(self, name=None):
+        """
+        删除DataSet中的collect_fn
+
+        :param name:
+        :return:
+        """
+        for _, dataset in self.datasets.items():
+            dataset.delete_collect_fn(name=name)
+
     def __repr__(self):
         _str = ''
         if len(self.datasets):

fastNLP/modules/encoder/lstm.py

@@ -24,7 +24,7 @@ class LSTM(nn.Module):
         """
         
         :param input_size:  输入 `x` 的特征维度
-        :param hidden_size: 隐状态 `h` 的特征维度.
+        :param hidden_size: 隐状态 `h` 的特征维度. 如果bidirectional为True，则输出的维度会是hidde_size*2
         :param num_layers: rnn的层数. Default: 1
         :param dropout: 层间dropout概率. Default: 0
         :param bidirectional: 若为 ``True``, 使用双向的RNN. Default: ``False``